Roses are the most important commercial ornamental plants grown for flowers, perfumery and nutraceutical compounds. Commercially cultivated roses (Rosa × hybrida L.) are complex interspecific hybrids probably derived from 8-10 wild species among the large diversity of 130-200 species in genus Rosa. Wild germplasm is a primary source of variability and plays a major role in improving existing varieties by broadening their genetic base. In the present investigation, we have utilized the previously identified SSR primers for studying the diversity among 148 selected rose genotypes, including wild species and cultivated varieties of Indian and exotic origin. A total of 88 alleles was scored using 30 polymorphic loci; they produced average 2.9±1 alleles per locus. Polymorphism information content (PIC) values for different SSR loci ranged from 0.08 to 0.8 with a mean value of 0.5±0.2. The neighbor-joining tree generated based on Nei’s (1978) genetic distance values grouped the population into three major clusters. Cluster-I and II consists of all modern rose cultivars (Rosa × hybrida L.) originated from India and cluster-III consists of all exotic cultivars, wild species and a few cultivars from India. STRUCTURE analysis based on microsatellite allelic data, partitioned the total rose genotypes into four different sub-populations with some individual genotypes having genomic admixture. Population subdivision estimates, FST between different subpopulations ranged from 0.01-0.15 indicates low to moderate level of divergence existing among the rose cultivars and germplasm. Population differentiation in rose cultivars and wild species corresponds to their geographical origin and lineages. Analysis of molecular variance (AMOVA) results revealed that 83.12 % of the variance was accounted for by within sub-groups followed by significant levels of variation among the populations (10.42%) and least variance (6.46%) was noticed among individuals within groups.
